Research On Key Parameter Optimization Method Of Aluminum Electrolysis Process Based On Data Drive

The aluminum electrolysis industry is a complex and high energy consumption process industry,and in the face of the current severe energy situation,energy saving and consumption reduction has become its primary goal.The key to the stable and efficient operation of the aluminum electrolysis production process is to ensure the normal operation of the aluminum electrolytic cell,which is also the basis for the enterprise to achieve economic benefits and energy saving and consumption reduction.The non-linear,multi-variable and large time lag characteristics of the aluminum electrolysis production process make it difficult to establish accurate mathematical models and achieve optimal process control by traditional methods.Therefore,taking the aluminum electrolysis process as the research background,the corresponding optimization measures are taken for different cell condition categories,which is important for stabilizing the aluminum electrolysis industrial production,reducing the aluminum production cost and achieving the purpose of energy saving and consumption reduction.In this thesis,a data-driven modeling approach and intelligent optimization algorithms are used to study the comprehensive evaluation of the electrolyzer state and the optimization of key parameters of the aluminum electrolysis process for the aluminum electrolysis production process,and certain research results have been achieved.The main research of this thesis is as follows.(1)Firstly,the structure of aluminum electrolyzer is described,the process mechanism of aluminum electrolysis production process and the key parameters of aluminum electrolysis are analyzed,and the control objectives of aluminum electrolysis production process are determined.The aluminum electrolysis data is normalized so as to eliminate the problem of weakened data analysis due to different magnitudes.(2)To solve the problem that the key parameters of aluminum electrolysis process are difficult to be measured accurately,an aluminum electrolysis key parameter prediction model based on improved least squares twin support vector regression is proposed.By combining mechanistic analysis and gray correlation analysis,the input variables of the prediction model are determined,the least squares twin support vector regression is used to establish the current efficiency and DC power consumption prediction models,and the quantum chaos salp swarm algorithm is used to optimize the model structure parameters.The simulation experimental results show that the established model has high prediction accuracy and can meet the needs of the actual soft measurement of aluminum electrolysis,which provides a basis for realizing multi-objective optimization of the aluminum electrolysis process.(3)When the aluminum electrolyzer is in different cell states,different control strategies are required in the cell control system.The electrolyte temperature is used as a performance indicator to evaluate the superiority of the cell condition by correlation parameter analysis of the electrolytic cell condition.In this thesis,an improved cell state judging model is proposed.The model uses convolutional neural network to extract deep features of aluminum electrolytic cell state,and then inputs the extracted deep features into the quantum chaos salp swarm algorithm to optimize the robust energy-based least squares twin support vector machine for cell state evaluation,and uses actual production data for verification.When an electrolyzer is in a poor cell,a retrospective analysis method is adopted and appropriate measures are given to restore it to a good or excellent cell.(4)When the state of the cell is judged as good or excellent cell,in order to achieve the optimization goal of energy saving and consumption reduction,a multi-objective optimal control model is established based on the prediction model of key parameters,with the maximum aluminum electrolysis current efficiency and the minimum DC power consumption as the optimization objectives and the aluminum electrolysis process parameters as the constraints.The model is solved using the third-generation non-dominated genetic algorithm.The simulation results show that the optimization strategy can ensure the stable operation of the aluminum electrolyzer and realize the optimized operation of aluminum electrolysis production,achieving a better energy saving and consumption reduction.